A novel enhanced bio-inspired harmony search algorithm for clustering

This paper presents a novel clustering algorithm based on harmony Search (HS) with features of Artificial Bee Colony (ABC) and K-means algorithms. Harmony search (HS) is a stochastic meta-heuristic optimization algorithm inspired from the improvisation process of musicians. However, HS fails to concentrate on the crucial factor of mutating the harmony vectors with better values in the improvisation step (where they are just pitch adjusted). The food source exploitation feature of ABC algorithm is applied to improve the members of the Harmony Memory based on their fitness values and hence improves the convergence rate of the Harmony Search method. This concept in combination with K-Means clustering algorithm leads to a new algorithm. To claim the superiority of the novel algorithm, its performance has been compared with the traditional HS, PSO, and K-means. The simulation results show that the proposed algorithm outperforms the other algorithms in terms of accuracy and convergence speed.